Free AWS Certified AI Business Strategist (AIB-C01) Practice Questions

Sample AIB-C01 Practice Questions

Browse all 10 free AWS Certified AI Business Strategist practice questions below.

1. A retail chain is reviewing three initiatives at a leadership meeting. The first is a demand forecasting system trained on five years of store sales history. The second is an assistant that drafts product descriptions from a short brief. The third is a rules engine that routes any refund under 50 USD for automatic approval. The chief operating officer asks which of the three is generative AI. Which response is correct?

   • The product description writer, because it produces new content in response to a prompt.
   • All three, because each one uses a model that learns patterns from company data.
   • The refund routing engine, because it makes decisions automatically without a person reviewing each case.
   • The demand forecasting system, because it was trained on historical data and produces numbers the company did not have before.

2. A regional bank has collected 30 candidate AI ideas from across the business during an internal campaign. The executive sponsor wants the first three initiatives to be chosen in a way the board will accept. The strategist has been asked to define the screening criteria before any idea is discussed on its merits. Which screening approach should be used?

   • Score each idea on business outcome, feasibility, and fit with the bank's stated objectives.
   • Rank ideas by how novel the underlying technology is, so the bank builds capability on the newest approaches first.
   • Choose the ideas that require the least data preparation so the first results arrive quickly.
   • Select the ideas submitted by the business units with the largest share of the bank's revenue.

3. A bank has deployed an AI model that recommends whether to approve small business loan applications. Applicants who are declined have a legal right to be told why. The model is accurate and well monitored, but the vendor describes its internal reasoning as proprietary and provides only a score with no factor breakdown. Which responsible AI dimension is most directly at risk?

   • Safety, because an incorrect decline could cause financial harm to a small business.
   • Privacy, because applicant financial data is being processed by a third party model.
   • Robustness, because the model may behave unpredictably on applications unlike those in its training data.
   • Explainability, because the bank must give declined applicants a reason for the decision.

4. A grocery chain's board has approved an ambitious AI agenda. The technology function reports that cloud infrastructure and data platforms are in place and that two data science teams are hired. The strategist has been asked to assess readiness before initiatives are launched, and finds that no executive owns the agenda, store managers have not been consulted, and there is no process for approving an AI system for use. What should the readiness assessment conclude?

   • The chain is ready, since infrastructure and technical talent are the binding constraints on AI delivery.
   • Readiness cannot be assessed until specific use cases are selected, since requirements differ by initiative.
   • The chain is not ready, since data quality has not been assessed and no readiness conclusion can be drawn without it.
   • The chain is technically ready but not organizationally ready on leadership, culture, and governance.

5. An insurance company wants to use AI to speed up claims triage. The data team reports that roughly 40 percent of claim records have a blank incident type field, and the free text adjuster notes use inconsistent abbreviations across three legacy systems that were merged after an acquisition. The program manager is preparing the business case for the steering committee. What is the most accurate business implication to present?

   • Missing fields are a technical concern for the data engineering team and do not change the business case.
   • Record volume matters more than record condition, so the project can proceed once the team confirms it has enough claims history.
   • Model outputs will carry the same gaps and inconsistencies, so data remediation belongs in the plan and budget before accuracy targets are set.
   • Free text adjuster notes are unstructured, so they should be excluded from the project scope.

6. A regional bank's credit team has been told that a new AI model outputs a probability rather than a yes or no decision. A senior manager asks why the system will not simply state whether an applicant should be approved, since that is what the team needs at the end of the process. Which explanation is correct?

   • The model returns a likelihood, and the bank sets the threshold that turns that likelihood into a decision.
   • The model is still in training, and it will produce firm decisions once enough applications have been processed.
   • Probability output indicates the model is uncertain, so the vendor should be asked for a more confident version.
   • The bank chose a generative model, and a predictive model would return the decision directly.

7. A B2B software company wants AI to improve sales performance. The sales director reports that win rates are healthy once a demo happens, but reps spend around 40 percent of their week researching accounts and writing outreach, and pipeline coverage is the constraint on revenue. Where should the AI investment be directed?

   • Sales forecasting, so leadership has a more accurate view of what will close each quarter.
   • Deal scoring, so reps prioritize the opportunities most likely to close.
   • Account research and outreach drafting, since that is where rep time is consumed.
   • Automated demo delivery, so more prospects can see the product without a rep present.

8. A gig economy platform uses a model to allocate delivery jobs to couriers. An internal review finds that couriers who decline jobs during school pickup hours receive fewer offers afterward, and that this pattern affects couriers with childcare responsibilities more than others. The model was never given any data about caring responsibilities. Which responsible AI dimension is most directly engaged?

   • Transparency, because couriers have not been told how job allocation decisions are made.
   • Privacy, because the model has inferred sensitive information about couriers' personal circumstances.
   • Fairness, because the allocation pattern produces systematically different outcomes for an identifiable group.
   • Robustness, because the model responds unpredictably to couriers who decline jobs.

9. A construction group scored well on a readiness assessment for data, infrastructure, and executive sponsorship. The one weak dimension was cultural preparedness: site managers described previous technology rollouts as things done to them, and two thirds said they expected the AI programme to be abandoned within a year like the last three initiatives. What does this finding mean for the programme?

   • It is a communication problem, and a launch campaign explaining the programme's benefits will address it.
   • Cultural readiness is a gating dimension, so the first initiative should visibly change that expectation.
   • It should be recorded as a risk and monitored, since culture changes slowly and cannot be addressed directly.
   • It reflects the previous initiatives rather than this one, so the programme should proceed on its own merits.

10. A logistics company is describing three systems to its board. The first flags shipments likely to arrive late based on ten years of delivery history. The second answers driver questions in natural language and drafts incident reports. The third checks whether a shipment weight exceeds a published limit and rejects it if so. The board wants each placed in the right category. Which categorization is correct?

    • All three are machine learning, since each one automates a judgment a person used to make.
    • The first is machine learning, the second is generative AI, and the third is rule-based automation.
    • The first two are both generative AI, since both produce an output the company did not previously have.
    • The first is generative AI, the second is machine learning, and the third is rule-based automation.